The subject of this grant, recurrence of Bipolar I (BP I) major depressive episode (MDE), is now recognized as a major mental health problem. Recurrent BP I MDE is a disorder with no satisfactory therapy, and its treatment remains a challenge to clinicians. To date, initial and long-term therapy of BP I MDE has been based on un-validated practice guidelines. These guidelines recommend limiting antidepressant drug (AD) use during initial therapy of BP I MDE, and completely avoiding AD use during long-term therapy. There is, however, no empirical evidence to suggest that mood stabilizer (MS) monotherapy is superior to combined MS plus AD therapy in preventing recurrent BP I MDE. Nor is there evidence to suggest that long-term MS plus AD therapy results in more manic switch episodes. We present evidence that AD-induced mania during long-term therapy of BP I MDE has been over-estimated, and that long-term use of MS plus AD therapy may be superior to MS therapy alone in preventing recurrent BP I MDE. In this application, we will ask: Does continuation therapy with combined lithium plus fluoxetine result in fewer MDE relapses and recurrences vs. lithium monotherapy? To answer this question, patients with BP I MDE will receive combined lithium plus fluoxetine therapy for 8 weeks. Responders who stay well for an additional 4 weeks of consolidation therapy will then be randomized to double-blind continuation therapy with either (i) combined lithium plus fluoxetine, or (ii) lithium alone (following fluoxetine taper and discontinuation) for an additional 50 weeks. We hypothesize that long-term lithium plus fluoxetine therapy will result in fewer MDE relapses and recurrences vs. lithium monotherapy. We will also ask: What is the relative safety, tolerability, and frequency of syndromal and sub-syndromal manic, hypomanic, and mixed state conversions during continuation treatment with combined lithium plus fluoxetine vs. lithium monotherapy? To answer this question, we will measure: the frequency, severity, and duration of syndromal and sub-syndromal manic, hypomanic, and mixed state conversions; frequency, severity, and duration of treatment-emergent adverse events; frequency of treatment discontinuation; time to onset of first syndromal and sub-syndromal conversion event; time to first treatment intervention of each syndromal and sub-syndromal conversion event; and, time to onset of increase in suicidal ideation event. We hypothesize that lithium plus fluoxetine therapy will result in a similar frequency of syndromal and sub-syndromal conversion events, and a similar frequency of treatment- emergent adverse events. We further hypothesize that lithium plus fluoxetine therapy will result in fewer suicide ideation events and fewer study discontinuations vs. lithium monotherapy. We believe that the results of this trial may have an important public health impact on the current practice guidelines for treating BP I MDE.